Multi-Instrument Error Coding Fault Diagnosis via Time Series Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for diagnosing error coding faults in petrochemical instruments are inadequate for complex, multi-instrument cooperation scenarios, leading to unreliable data and potential safety hazards in petrochemical production.
Innovation Solution
A method and system that acquires sampling data series from multiple instruments, determines the type of data series as either weakly stationary or non-stationary, and uses semiorder relation analysis and sliding outlier-tolerant filtering to perform error diagnosis, eliminating time-varying components and ensuring accurate fault detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional fault diagnosis methods are used for petrochemical instruments, then the diagnosis process is simple, but the accuracy and reliability of fault detection are insufficient for complex multi-instrument scenarios
Solution Approach 1:
The patent segments the fault diagnosis process into distinct modules: data acquisition module, time series type judgment module, and differentiated diagnosis modules. For weakly stationary time series, it applies semiorder relation analysis; for non-stationary time series, it applies sliding outlier-tolerant filtering. This segmentation allows complex multi-instrument fault diagnosis to be handled through systematic, modular processing steps, improving accuracy without overwhelming complexity
Solution Approach 2:
The patent changes the parameter of time series stationarity (weakly stationary vs. non-stationary) to determine which diagnosis method to apply. By identifying the stationarity parameter of the sampling data series, the system adapts its diagnosis approach accordingly, enabling accurate fault detection across diverse instrument conditions while maintaining a structured diagnostic framework
2Adaptability or versatility
If direct diagnosis based on mechanism analysis is used, then the diagnosis method is intuitive and simple, but it cannot handle complex cases with multi-factor coupling and multi-instrument cooperative work
Solution Approach 1:
The patent creates a universal diagnosis system that can handle both weakly stationary and non-stationary time series through a single integrated framework. The system first judges the time series type and then applies the appropriate diagnosis method, making it adaptable to various instrument fault scenarios including multi-factor coupling and multi-instrument cooperative work, while maintaining systematic complexity management through standardized processing flows
3Reliability
If instruments are used for long-term service in petrochemical environments, then the production monitoring coverage is comprehensive, but the instruments are subject to oil and gas corrosion and develop faults
Solution Approach 1:
The patent implements continuous monitoring and feedback through the diagnosis system that constantly analyzes sampling data from multiple instruments. By detecting faults early through systematic analysis of time series data and providing feedback on instrument status, the system enables timely maintenance decisions, ensuring reliable production monitoring while mitigating the effects of corrosion and environmental degradation on long-term instrument service
Data Source
AI summary
A method and system for diagnosis of error coding faults from multiple instruments are provided. The method includes acquiring sampling data series of a combination of instruments in a petrochemical process, determining a type of the sampling data series, and performing error diagnosis according to the type. The present disclosure can solve the error coding problem in a multi-instrument cooperation mode and provide safe and reliable data guarantee for safe and efficient petrochemical production.